Budget-aware IDS for ICS/SCADA
Cost-aware feature acquisition and strict FPR calibration for real-time detection in constrained environments.
Open to ML/AI Research roles • Security ML • Anomaly Detection • CPS/ICS
Project builds that show depth, rigor, and deployability.
A collection of implemented systems, experiments, and research prototypes—plus what I’m building next. Most completed projects link to GitHub for code and reproducibility.
—
Public repos
—
Stars
—
Followers
What employers care about
Proof of execution: clean code, disciplined evaluation, and deployable systems.
- Real-time ML systems under constraints (FPR/latency)
- Strong experimental discipline (ablations, calibration)
- Clear technical writing and documentation
- Reproducible code on GitHub
Featured Projects
High-signal work that maps directly to ML/security research roles.
Smart water metering: event vs attack detection
Robust residual modeling and partition-aware monitoring to distinguish natural events from false-data attacks.
Context-aware RL feature selection
Learning a sparse, budget-feasible feature mask that balances detection accuracy and acquisition cost.
GitHub Projects
Automatically pulled from GitHub. Filter by keyword and sort by what matters.
Loading repositories…
Upcoming Work
What I’m actively building next (and why it matters for deployment and impact).
Research-to-production IDS package
Packaging detection + calibration + evaluation into a clean, reusable repo with reproducible configs and CI.
Latency & FPR guaranteed evaluation suite
A standardized evaluation harness for security ML: latency distribution reporting and strict FPR control.
Public write-ups + repo docs
Short, clear technical notes per project—threat model, method, results, and “how to run”.
Want to collaborate or hire?
Reach out and I’ll share the most relevant repos and results for your role.